Infectious disease with antibiotic resistance is considered the most imperative global health problem. Early detection of infection is necessary to institute antibiotic therapy as soon as possible. The spatial localization of infection within the body often dictates treatment. Spatial localization is also important in diagnosis, for example, in determining whether a painful joint prosthesis is infected or merely loose or whether a vascular graft in a patient presenting with fever is indeed infected and needs excision. We have developed a specific technique for the imaging of infection that is applicable to most bacteria and many viruses. The method uses radiolabeled nucleoside analogs that are substrates of bacterial and viral thymidine kinases (TKs). Upon entering the cell, these compounds are phosphorylated, trapped and concentrated within cells harboring the TK enzyme. We have successfully demonstrated the ability to image a variety of bacterial species both in animal models and in a limited clinical study. We have also been able to image tumors infected with gamma herpesviruses by virtue of the viral TK. The current state-of-the-art for imaging infection involves a laborious process of isolating and tagging leukocytes from a patient and re-administering them. That technique often provides false positives (sterile inflammation) and is cumbersome and inefficient. The emerging clinical replacement for tagged leukocytes, particularly for chronic infections of the axial skeleton, is positron emission tomography with [18F]fluorodeoxyglucose (FDG-PET). Our technique radiolabels and images bacteria directly. We anticipate that our technique will provide a viable alternative to the way infection is imaged, but have only tested it in one, limited, clinical application: musculoskeletal infection. In this proposal we will further study musculoskeletal infection, comparing this new method to FDG-PET, while attempting to determine the sensitivity and especially the specificity of our technique. To determine specificity we will image a cohort of patients with rheumatoid arthritis, a sterile form of joint inflammation. We will also extend our method to one new clinical indication by studying a small cohort of patients with possible pulmonary infection (Aim 2). The imaging agent that we have been using to date has been 2'-fluoro-2'-deoxy-1-2-D-arabinofuranosyl-5- [124I]iodouracil ([124I]FIAU) for PET combined with computed tomography (PET/CT). In addition to performing the first two aims using [124I]FIAU, the third aim will involve synthesis of the corresponding material labeled with 18F in an automated fashion to enable an eventual transition to [18F]FIAU. We may ultimately use the 18F- labeled material in future studies. We believe that this project will help to establish this nucleoside-based imaging technique as a new alternative for imaging infection.